This is a revised version of an application to study the molecular genetics of H. pylori (Hp). Dr. Berg took the critique of the previous version so much to heart that he has prepared what is in essence an entirely new application. Therefore, he has not responded in detail to the previous critique and no more need be said about it. The present application is focussed on emerging antibiotic resistance in Hp and on the possibility of extensive genetic exchange within the species. Dr. Berg notes that Hp shows a remarkable genotypic diversity with respect to fingerprinting markers as well as specific proteins such as those implicated in virulence. The molecular basis for this diversity is, however, unknown, though it has been observed that Hp isolates are frequently transformable, often highly so, and could engage in DNA exchange in vivo. As might have been expected, resistance to clathromycin (Cla) and metronidazole (Met), two widely used drugs, has been on the rise, especially in the third world, where these drugs are used indiscriminately. Neither the mechanisms of resistance nor the responsible genes have been identified to date. Dr. Berg has two specific aims, namely to identify and characterize the genes for Cla and Met resistances and to analyze genetic recombination among Hp strains in vivo. In preliminary studies, he has developed the RAPD and PCR-RFLP methods for genotyping of Hp,and shown that these methods are more sensitive than MLEE, and that plasmid variability does not contribute disproportionately to typing specificity. He also showed that RAPD was useful for pedigree analysis of laboratory K12 strains and that observed variations were often at or near the replication terminus, possibly owing to decay of the replication complex at the terminus in strains stored on stabs. With respect to Hp, Dr. Berg has demonstrated infection with multiple Hp strains and confirmed joint multiple infection in a gnotobioticpiglet model - which appears to be a very good model for Hp infection. Finally, he has constructed an ordered cosmid library of 68 units with 3 gaps, and found the order of genes to be quite different from that in a different strain mapped by PFGE.